Productivity characteristics of high-performing commercial swine breeding farms

Study on the influence of different production factors on PSY and its correlation

Background

Finding out the key reproductive performance factors, affecting piglets weaned per sow per year (PSY) can improve the production efficiency and profitability of pig farms. The objective was to understand the actual distribution of different production factors and PSY of breeding pig farms, analyze the correlation to find the main production factors affecting PSY, and formulating a Production Efficiency Improvement Plan in practice. Data included 603 breeding pig farms from September 28, 2020 to September 26, 2021. Regression analysis was used to evaluate the relationship between PSY and key production factors, and the characteristics of total pig farms versus high performance (HP) pig farms (the production performance was in the top 10%) or top 5% pig farms were compared. Spearman’s rank correlation coefficient was used to analyze the correlation between production factors and find the factors related to PSY. Non-linear support vector regression (NL-SVR) was used to analyze the personalized PSY improvement through a various change of the four key factors.

Results

The median distribution of 15 production factors and PSY in total pig farms were different from those of HP farms. All of data were distributed nonlinearly. Mating rate within 7 days after weaning (MR7DW), farrowing rate (FR), number of piglets born alive per litter (PBAL) and number of weaned piglets per litter (WPL) were moderately correlated with PSY, and the correlation coefficients were 0.5058, 0.4427, 0.3929 and 0.3839, respectively. When the four factors in NL-SVR changed in medium (0.5 piglet or 5%) or high level (1.0 piglet or 10%), PSY can be increased by more than 0.5.

Conclusion

NL-SVR model can be used to analyze the impact of changes in key production factors on PSY. By taking measures to improve MR7DW, FR, PBAL and WPL, it may effectively improve the current PSY and fully develop the reproductive potential of sows.

Utilization and reproductive performance of gilts in large-scale pig farming system with different production levels in China: a descriptive study

Background

This study was to investigate the utilization and reproductive performance of gilts in large-scale pig farms. Data of this descriptive study included 169,013 gilts of 1540 gilts’ batches on 105 large-scale pig farms from April 2020 to March 2021. According to the upper and lower 25th percentiles of piglets weaned per sow per year (PSY) during the research stage, pig farms were divided into three productivity groups: high-performing (HP), intermediate-performing (IP) and low-performing (LP) farms. On the basis of breeds, LP (LP-Total) farms was further divided into LP-breeding pig (LP-BP) and LP-commercial pig (LP-CP) groups. Average utilization, estrus and first mating data was collected from a total of 1540 gilts’ batches. The age-related factors (introduction age, age at first estrus and age at first mating) and litter production (total number of piglets, number of piglets born alive and number of weaned piglets, as well as their proportion distribution) among HP and LP groups were compared. The litter production in different age groups were also analyzed.

Results

The introduction age, mortality and culling rate of HP farms were lower compared with LP farms. Total number of piglets per litter, number of piglets born alive per litter and number of weaned piglets per litter in HP farms were significantly more than those of LP groups, respectively. The proportion distribution peaks of litter production in HP farms were shifted about two more than those in LP groups, respectively; and the proportion of low litter production (eight per litter or less) was lower than that in LP groups. The results of different age groups showed that total number of piglets per litter and number of piglets born alive per litter in 220–279 d were the most, while that of 370 d was the least.

Conclusions

The overall utilization and reproductive performance of gilts in HP farms was better than those of LP farms. The difference in utilization was reflected in introduction source, culling rate and mortality. While the age at first estrus and first mating, breeds and litter production were the main differences for reproductive performance.

Farm data analysis for lifetime performance components of sows and their predictors in breeding herds

Our objectives in this review are 1) to define the four components of sow lifetime performance, 2) to organize the four components and other key measures in a lifetime performance tree, and 3) to compile information about sow and herd-level predictors for sow lifetime performance that can help producers or veterinarians improve their decision making. First, we defined the four components of sow lifetime performance: lifetime efficiency, sow longevity, fertility and prolificacy. We propose that lifetime efficiency should be measured as annualized piglets weaned or annualized piglets born alive which is an integrated measure for sow lifetime performance, whereas longevity should be measured as sow life days and herd-life days which are the number of days from birth to removal and the number of days from date of first-mating to removal, respectively. We also propose that fertility should be measured as lifetime non-productive days, whereas prolificacy should be measured as lifetime pigs born alive. Second, we propose two lifetime performance trees for annualized piglets weaned and annualized piglets born alive, respectively, and show inter-relationships between the four components of the lifetime performance in these trees. Third, we describe sow and herd-level predictors for high lifetime performance of sows. An example of a sow-level predictor is that gilts with lower age at first-mating are associated with higher lifetime performance in all four components. Other examples are that no re-service in parity 0 and shorter weaning-to-first-mating interval in parity 1 are associated with higher fertility, whereas more piglets born in parity 1 is associated with higher prolificacy. It appears that fertility and prolificacy are independent each other. Furthermore, sows with high prolificacy and high fertility are more likely to have high longevity and high efficiency. Also, an increased number of stillborn piglets indicates that sows have farrowing difficulty or a herd health problem. Regarding herd-level predictors, large herd size is associated with higher efficiency. Also, herd-level predictors can interact with sow level predictors for sow lifetime performance. For example, sow longevity decreases more in large herds than small-to-mid herds, whereas gilt age at first-mating increases. So, it appears that herd size alters the impact of delayed gilt age at first-mating on sow longevity. Increased knowledge of these four components of sow lifetime performance and their predictors should help producers and veterinarians maximize a sow's potential and optimize her lifetime productivity in breeding herds.

High-performing farms exploit reproductive potential of high and low prolific sows better than low-performing farms

Background

Our objective was to examine the impact of farm effects and sow potential on various aspects of sow performance. We examined the interaction between sow prolificacy groups categorized at parity 1 and farm productivity groups for reproductive performance across parities, and lifetime performance. Data included 419,290 service records of 85,096 sows, on 98 Spanish farms, from first-service as gilts to removal, that were served between 2008 and 2013. Farms were categorized into three productivity groups based on the upper and lower 25th percentiles of the farm means of annualized lifetime piglets weaned per sow over the 6 years: high-performing (HP), intermediate-performing (IP), and low-performing (LP) farms. Also, parity 1 sows were categorized into three groups based on the upper and lower 10th percentiles of piglets born alive (PBA) as follows: 15 piglets or more (H-prolific), 8 to 14 piglets, and 7 piglets or fewer (L-prolific). The farm groups represent farm effects, whereas the sow groups represent sow potential. Linear mixed effects models were performed with factorial arrangements and repeated measures.

Results

Mean parity at removal (4.8 ± 0.01) was not associated with three farm productivity groups (P = 0.43). However, HP farms had 7.7% higher farrowing rates than LP farms (P <  0.05). As a result, H-prolific and L-prolific sows on HP farms had 29.7 and 30.7 fewer non-productive days during lifetime than the respective sows on LP farms (P <  0.05). Furthermore, the H-prolific and L-prolific sows on HP farms had 4.9 and 6.2 more annualized piglets weaned than respective H-prolific and L-prolific sows on LP farms (P <  0.05), which was achieved by giving birth to 0.8–1.0 and 1.4–1.7 more PBA per litter, respectively, than on HP farms during parities 2–6 (P <  0.05). During the first parity, HP farms had 18.8% H-prolific sows compared to 6.2% on LP farms.

Conclusion

Farm effects substantially affected lifetime performance of sows. Higher lifetime productivity of sows on HP farms was achieved by higher farrowing rate, fewer non-productive days, more PBA and more piglets weaned per sow, regardless of prolific category of the sows.

Factors for improving reproductive performance of sows and herd productivity in commercial breeding herds

We review critical factors associated with reproductive performance of female breeding pigs, their lifetime performance and herd productivity in commercial herds. The factors include both sow-level and herd-level factors. High risk sow-level groups for decreasing reproductive performance of female pigs are low or high parity, increased outdoor temperature, decreased lactation feed intake, single inseminations, increased lactation length, prolonged weaning-to-first-mating interval, low birth weight or low preweaning growth rate, a few pigs born alive at parity 1, an increased number of stillborn piglets, foster-in or nurse sow practices and low or high age at first-mating. Also, returned female pigs are at risk having a recurrence of returning to estrus, and female pigs around farrowing are more at risk of dying. Herd-level risk groups include female pigs being fed in low efficiency breeding herds, late insemination timing, high within-herd variability in pig flow, limited numbers of farrowing spaces and fluctuating age structure. To maximize the reproductive potential of female pigs, producers are recommended to closely monitor females in these high-risk groups and improve herd management. Additionally, herd management and performance measurements in high-performing herds should be targeted.

Application of LogitBoost Classifier for Traceability Using SNP Chip Data

Consumer attention to food safety has increased rapidly due to animal-related diseases; therefore, it is important to identify their places of origin (POO) for safety purposes. However, only a few studies have addressed this issue and focused on machine learning-based approaches. In the present study, classification analyses were performed using a customized SNP chip for POO prediction. To accomplish this, 4,122 pigs originating from 104 farms were genotyped using the SNP chip. Several factors were considered to establish the best prediction model based on these data. We also assessed the applicability of the suggested model using a kinship coefficient-filtering approach. Our results showed that the LogitBoost-based prediction model outperformed other classifiers in terms of classification performance under most conditions. Specifically, a greater level of accuracy was observed when a higher kinship-based cutoff was employed. These results demonstrated the applicability of a machine learning-based approach using SNP chip data for practical traceability.

Swine herds achieve high performance by culling low lifetime efficiency sows in early parity

Sow lifetime performance and by-parity performance were analyzed using a 3 by 3 factorial design, comprising 3 herd productivity groups and 3 sow efficiency groups. Data was obtained from 101 Japanese herds, totaling 173,526 parity records of 34,929 sows, for the years 2001 to 2006. Sows were categorized into 3 groups based on the lower and upper 25th percentiles of the annualized lifetime pigs born alive: low lifetime efficiency sows (LE sows), intermediate lifetime efficiency sows or high lifetime efficiency sows. Herds were grouped on the basis of the upper and lower 25th percentiles of pigs weaned per mated female per year, averaged over 6 years: high-, intermediate- or low-performing herds. Mixed-effects models were used for comparisons. LE sows in high-performing herds had 57.8 fewer lifetime nonproductive days and 0.5 earlier parity at removal than those in low-performing herds (P<0.05). The number of pigs born alive of LE sows continuously decreased from parity 1 to 5, whereas those of high lifetime efficiency sows gradually increased from parity 1 to 4 before decreasing up to parity ≥ 6 (P<0.05). In conclusion, the LE sows have a performance pattern of decreasing number of pigs born alive across parity. The present study also indicates that high-performing herds culled potential LE sows earlier than the other herds.

Double and triple matings associated with reproductive performance in first-serviced and reserviced female pigs in commercial herds

The objective of this study was to investigate associations of the number of matings and services with reproductive performance in high-performing and ordinary herds. The data included 113,265 service and 92,248 farrowing records in 117 herds. A service included single or more matings of a female pig (female) during a 10-day estrus period. Two herd groups were built on the basis of the upper 25th percentile of pigs weaned per mated female per year: high-performing (> or = 22.8 pigs) and ordinary herds. Mixed-effects models were used to analyze reproductive performance. Relative frequencies (%) of single, double and triple or more matings were 3.4, 27.4, and 69.2% in high-performing herds, respectively, and were 4.6, 59.3 and 36.1% in ordinary herds, respectively. Percentages of reserviced females in high-performing and ordinary herds were 7.3 and 13.0%, respectively. Triple or more-mated (TM) gilts had 3.5% higher farrowing rates than double-mated (DM) gilts (P<0.01), but similar pigs born alive (PBA) to DM gilts in the first service group in both the herd groups. In the first service group, TM sows had 0.8% higher farrowing rates and 0.2 more PBA than DM sows in high-performing herds (P<0.01). In the reservice group, TM gilts and TM sows had farrowing rate similar to DM gilts and DM sows in high-performing herds. In conclusion, performing triple matings was a better practice for first-serviced females than performing double matings. Double matings may be sufficient for reserviced females.

By-parity nonproductive days and mating and culling measurements of female pigs in commercial breeding herds

The objectives of this study were to determine by-parity nonproductive female days (NPD or NPDs) and mating and culling measurements, to determine correlations between by-parity NPDs, mating and culling measurements and herd productivity measurements, and to compare by-parity NPDs between three herd groups (105 herds) with differing reproductive productivities. NPD was defined as the number of days when mated females were neither gestating nor lactating. Correlation analysis and mixed-effects models were performed. On the basis of the 25th and 75th percentiles of pigs weaned per mated female per year, three herd groups were formed: high-, intermediate-, and low-performing herds. The mean NPD of 105 breeding herds (mean +/- SEM) was 52.7 +/- 1.6 days. The NPDs in parities 1, 6 and > or = 7 were higher than those in parities 0, 2, 3 and 4 (P<0.05). High-performing herds had a higher farrowing percentage and lower percentage of reserviced females than low-performing herds (P<0.05). Lower by-parity NPDs were correlated with lower percentages of reserviced females, higher farrowing percentages and lower culling rates from parities 1 to 5 (P<0.05). High-performing herds had NPDs that were > 25 days lower in parities 0 to 3 than low-performing herds (P<0.05). High-performing herds had lower culling rates in parities 2 to 5 and higher culling rates in parities 6 and > or = 7 than low-performing herds (P<0.05). The present study indicates that monitoring the by-parity NPD and mating and culling measurements is a good tool for improvement of herd productivity.

Lactational performance for improving postweaning reproductive performance and lifetime performance on swine commercial farms

We studied the associations of the weaning litter weight (WLWt) and number of pigs weaned (PW) with measurements of postweaning reproductive performance and examined the repeatability of WLWt, PW and average pig weight at weaning (PIGWt) on commercial swine farms. This study spanned 6 years and was conducted using 57,611 weaning records from 11,574 sows born in 1999 on 92 farms. Variance components analysis was used to determine the repeatability of measurements of lactational performance. Mixed-effects models were used to analyze the associations of measurements of lactational performance with farrowing rate and weaning-to-first-mating interval. The values for repeatability of WLWt, PIGWt and PW were 0.31, 0.34 and 0.17, respectively. No differences in weaning-to-first-mating intervals were found among the five PW groups (< or = 6, 7 to 8, 9 to 10, 11 and 12 to 14 pigs) or among the three WLWt groups (< or = 48.0, 48.0 to 69.0 and > or = 69.0 kg). Sows with 12 to 14 PW had farrowing rates similar to those with 9 to 11 PW. Sows with a WLWt > or = 69.0 kg had the highest farrowing rate (P<0.01). However, sows with 11 PW had an approximately 100 to 200 g lighter PIGWt than those with 4 to 10 PW (P<0.01). This study suggests that increased WLWt and PW do not impair postweaning reproductive performance, but instead decrease PIGWt.

Technical note: High-performing swine herds improved their reproductive performance differently from ordinary herds for five years1

The objectives of this study were to determine changes in herd productivity and the performance of female pigs over time in commercial swine herds. Annual measurement data from 1999 to 2003 were obtained from the record files of 113 herds in Japan. Two groups were formed according to the 25th percentile of pigs weaned/mated females per year (PWMFY) in 2003; the 2 groups were high-performing herds (those constituting the top 25%) and the remaining ordinary herds. The effects of group based on PWMFY in 2003, year, and the group x year interaction on repeated measures between 1999 and 2003 were analyzed by using mixed-effects models. A regression analysis was also used to compare key measurements in productivity between the 2 groups, with years as a continuous variable. Variance components were obtained to determine herd repeatability of PWMFY for the 2 herd groups. The average female inventory increased from 290 +/- 31 to 355 +/- 42 females for these 5 yr. The PWMFY also changed from 20.9 +/- 0.21 to 21.2 +/- 0.30 pigs. An interaction between year and group was detected (P < 0.05) for PWMFY. In the regression comparison, high-performing herds increased their PWMFY by 0.31 +/- 0.09 pigs each year, whereas the ordinary herds did not increase. The number of pigs weaned per sow increased by 0.07 +/- 0.02 pigs each year in high-performing herds and increased by 0.03 +/- 0.01 pigs each year in ordinary herds. In high-performing herds, for each year, the percentage of sows mated by 7 d after weaning increased by 0.92 +/- 0.25%, the percentage of reserviced females decreased by 0.63 +/- 0.26%, and culling rate increased by 1.53 +/- 0.50%. Repeatability of PWMFY for high-performing herds and ordinary herds was 28.8 and 54.0%, respectively. This study shows that productivity in high-performing herds was improved compared with that of ordinary herds.

Six component intervals of nonproductive days by breeding-female pigs on commercial farms

Of 105 swine herds using a production record system for breeding female pigs, 95 farms were used to analyze nonproductive female days (NPD), the six component intervals of NPD, and related measurements. The NPD was defined as the days when mated gilts and sows were neither gestating nor lactating, and it was calculated by summing the six component intervals in the average mated female inventory. The mean NPD was 57.9 d (SD = 20.5), and the proportions of six component intervals of gilt first-mating-to-pregnancy interval, gilt first-mating-to-culling interval, unmated weaning-to-culling interval, weaning-to-first-mating interval, sow first-mating-to-pregnancy interval, and sow first-mating-to-culling interval were 9.24, 7.82, 6.85, 27.9, 18.9, and 29.3%, respectively. Farms in the upper 25th percentile of the ranking for number of pigs weaned.mated female(-1).yr(-1) were designated as 25 high-performing farms. The remaining farms were designated as an ordinary farm group for comparisons. High-performing farms had 21.1 d fewer NPD, and five of the six component intervals were lower compared with the ordinary farms (P < 0.05). Regression analyses indicated that the number of litters.mated female(-1).yr(-1) increased by 0.07 in both farm groups as NPD decreased every 10 d. Fewer NPD were correlated with a higher percentage of multiple matings during estrus (P < 0.05) but were not correlated with removal risk and replacement risk in both farm groups. The average parity of culled females was negatively correlated with NPD in the ordinary farm group, and the average farrowed parity was positively correlated with NPD in the high-performing farm group (P < 0.01). Decreasing each component interval of the NPD six components is critical to increasing herd productivity. A high percentage of multiple matings during estrus and appropriate culling management may be key factors to decrease NPD.

Longevity and efficiency associated with age structures of female pigs and herd management in commercial breeding herds

Annual performance measurements, age structures of female pig inventories, and by-parity culling rates were abstracted from data files of 110 herds that participated in a data-share program in Japan. Parity at culling was used as a prime measurement of longevity, whereas pigs weaned x mated female(-1) x year(-1) (PWMFY) was a prime measurement of reproductive efficiency. High or low longevity herds were based on the greatest 50% of the herds or the remaining herds ranked by parity at culling, whereas high or low reproductive efficiency herds were grouped according to the greatest 50% of the herds or the remaining herds ranked by PWMFY. Measurements were analyzed as a 2 x 2 factorial arrangement, using the main effects of the 2 herd groups of longevity (high or low) and reproductive efficiency (high or low). Means of parity at culling and PWMFY were 4.6 (SD = 0.82) and 21.2 (SD = 3.02), respectively. The high longevity group had 1.27 greater parities at culling than the low longevity group (P < 0.05), but no differences between the high and low longevity groups were found in PWMFY (P = 0.21). No differences between the high and low efficiency groups were found in parity at culling (P = 0.50). No interactions between the longevity and efficiency groups were found on any longevity or efficiency measurement (P > 0.20). In herd management, the percentage of reserviced females and the percentage of multiple matings were associated with the longevity group and the efficiency group (P < 0.05). The high longevity group had lower culling rates in parity 0 to 6 than the low longevity group (P < 0.05), whereas no differences between the low and high efficiency groups were found in culling rates in parity 0 to 2 (P > 0.20). This study suggests that measures to achieve longevity and high reproductive efficiency in breeding herds do not conflict and that high reproductive efficiency and high longevity can be achieved.

Within-farm variability in age structure of breeding-female pigs and reproductive performance on commercial swine breeding farms

This study investigated relationships between herd age structure and herd productivity in breeding herds; it also investigated a pattern in parity proportions of females over 2 years and its relationship with herd productivity in commercial swine herds. This study was based on data from 148 commercial farms in North America stored in the swine database program at the University of Minnesota. The primary selection criterion was fluctuations in breeding-female pig (female) inventories over a 2-year interval. Productivity measurements and parity proportions of females were extracted from the database. A 24-month time-plot in proportions of Parity 0 and Parities 3-5 females (mid-parity) was charted for each farm. Using these charts, a change in proportions of Parity 0 and mid-parity for each farm was categorized into patterns: FLUCTUATE (Parity 0 and mid-parity proportion lines crossed) or STABLE (the two proportion lines never crossed). Higher proportions of mid-parity sows were correlated with greater pigs weaned per female per year (PWFY; P < 0.01). Farms with a FLUCTUATE (73% of the 148 farms) pattern had lower PWFY than those with a STABLE pattern (P < 0.01). The STABLE farms had higher proportions of mid-parity sows, higher parity at culling, higher frequency of gilt deliveries per year, and lower replacement rate than the FLUCTUATE farms (P < 0.01). In conclusion, maintaining stable subpopulations with mid-parity and Parity 0 are recommended to optimize herd productivity.

Re-Serviced Females on Commercial Swine Breeding Farms

The objectives of this study were to observe subsequent reproductive performance of re-serviced females by the number of services within a parity, to measure mean days to re-service and culling intervals, to determine lifetime performance in re-serviced gilts, and to investigate re-serviced females across parities on commercial farms. Reproduction records on 539 U.S.A. farms were used to observe re-serviced females by the number of service groups at the herd level. Farrowing rate decreases by approximately 10%, and re-service occurrence increases by approximately 5% for each increase in the number of services increase within a parity group (P<0.05). Only in parity 0 to 2 groups, average pigs born alive at subsequent farrowing in the second or later service groups were greater than in the first service group (P<0.05), but in parity >or=3, the third or later service groups produced fewer pigs born alive than the other service groups (P<0.05). Lifetime performance and re-service events were observed in 39945 individual females on the 149 selected farms that had complete 5-year records. Means of days to re-service, first-mating-to-culling intervals in gilts and weaning-to-culling intervals in sows were 46.3 days, 95.2 days, and 48.2 days, respectively. Re-serviced gilts had longer NPD (>50 days), a lower parity at culling (>0.5) and fewer lifetime pigs born alive (>2 pigs) than non-return gilts (P<0.05), but no difference in average pigs born alive per parity was found between re-serviced gilt groups and non-return gilts. Of 19677 re-serviced females, 35.6% had two or more re-services across parities in pig life, 10.6% had 3 or more re-services, and 1.95% had four or more re-services. Accurate estrus detection with a boar and improved mating techniques on re-serviced females are suggested to improve herd productivity.

Reproductive productivity measurements in Japanese swine breeding herds

To set productivity standards and targets, and investigate inter-relationships between key measurements in swine breeding herds, farm productivity measurements were analyzed on 87 Japanese commercial farms in 14 prefectures. The 87 herds were ranked on the basis of number of pigs weaned per mated female per year (PWMFY), and 23 herds in the upper 25th percentile of this ranking were designated as high-performing farms. Productivity measurements on the high-performing farms were compared with values for the remaining farms. The high-performing farms had shorter farrowing intervals, greater litters per mated female per year (LMFY), greater pigs weaned per sow (PWS), and greater mean parity of culled sows than the remaining farms (P<0.01). No difference in lactation duration was found between the two groups (P>0.10). For both farm groups, correlations of key reproductive measurements were determined. Lactation duration was not correlated with LMFY, PWS and PWMFY on the high-performing farms, while short lactation duration was correlated with greater LMFY and PWMFY on the remaining farms (P<0.01). In contrast to lactation duration, farrowing interval was correlated with PWS on the high-performing farms, but not on the remaining farms. Mean parity of culled sows were correlated to PWS and pigs born alive per sow on the high performing farms, but not with any measurements on the remaining farms. These results suggest that high-performing farms have used different herd management from the remaining farms.